The mapping of trace elements on fine grids from dense data on covariates from remote sensors and high-resolution environmental predictors

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00027049%3A_____%2F22%3AN0000058" target="_blank" >RIV/00027049:_____/22:N0000058 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

The mapping of trace elements on fine grids from dense data on covariates from remote sensors and high-resolution environmental predictors

Popis výsledku v původním jazyce

The distribution of trace elements in the surface soils is complex and reflects the geochemistry of the original geological substrate modified by a variety of environmental and human-induced factors. Since the covariate datasets representing these factors are usually in a much finer resolution than the geochemical surveys, the auxiliary variables were employed for the predictive mapping of As, Cd, Ni, and Pb using a quantile random forest model. Taking into account the multi-conditional nature of topsoil geochemistry, various remote sensed or high-resolution covariates were selected for predictive mapping within the study in the Czech Republic supported by the Technology Agency of the Czech Republic – project No. SS03010364. Geological variation was conceptualised using the airborne geophysical (gravimetric, radiometric, and magnetometric) data together with local effects represented by the intensity of mineral exploration from an inventory of mining dumps. Soil conditions for the binding capacities were represented by DEM derivatives and predictions for organic carbon and clay contents from dense legacy data. Finally, the precipitation and deposition rates from continuous ground measurements, and night-time lights data covered the natural and human-induced distributional effects. Moreover, the model was tested for an increase of prediction capability when using a mosaic of bare soils from Sentinel-2 satellite data or the Euclidean buffer distance maps to better capture the spatial dependencies. The relative importance of covariates and cross-validated metrics varied across the range of elements which hinted at the controls of their distributions or the effects of underrepresented processes within the suite of auxiliary variables.

Název v anglickém jazyce

The mapping of trace elements on fine grids from dense data on covariates from remote sensors and high-resolution environmental predictors

Popis výsledku anglicky

The distribution of trace elements in the surface soils is complex and reflects the geochemistry of the original geological substrate modified by a variety of environmental and human-induced factors. Since the covariate datasets representing these factors are usually in a much finer resolution than the geochemical surveys, the auxiliary variables were employed for the predictive mapping of As, Cd, Ni, and Pb using a quantile random forest model. Taking into account the multi-conditional nature of topsoil geochemistry, various remote sensed or high-resolution covariates were selected for predictive mapping within the study in the Czech Republic supported by the Technology Agency of the Czech Republic – project No. SS03010364. Geological variation was conceptualised using the airborne geophysical (gravimetric, radiometric, and magnetometric) data together with local effects represented by the intensity of mineral exploration from an inventory of mining dumps. Soil conditions for the binding capacities were represented by DEM derivatives and predictions for organic carbon and clay contents from dense legacy data. Finally, the precipitation and deposition rates from continuous ground measurements, and night-time lights data covered the natural and human-induced distributional effects. Moreover, the model was tested for an increase of prediction capability when using a mosaic of bare soils from Sentinel-2 satellite data or the Euclidean buffer distance maps to better capture the spatial dependencies. The relative importance of covariates and cross-validated metrics varied across the range of elements which hinted at the controls of their distributions or the effects of underrepresented processes within the suite of auxiliary variables.

Druh

O - Ostatní výsledky

CEP obor

—

OECD FORD obor

40104 - Soil science

Projekt

<a href="/cs/project/SS03010364" target="_blank" >SS03010364: Systém na podporu rozhodování při hodnocení kvality půdy z hlediska obsahu rizikových látek v zemědělských půdách České republiky</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2022

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů